Normalized Texture Motifs and Their Application to Statistical Object Modeling

Abstract

Abstract

Graph Partitioning Active Contours for Knowledge-Based Geo-Spatial Segmentation

“…We model a given geo-spatial object class by statistically learning [23,24] the multiple textures that characterize the class. This model captures the wide variation of visual features within the class.…”

“…To achieve rotation invariance, we construct a Gaussian mixture model [24] that learns the equivalence between different versions of feature vectors caused by rotation. The conditional probability of a feature vector x, given that it is generated from cluster j and its orientation index is r, is written as…”